This paper examines the middle ear of fossil living animals in terms of the homologies which have been drawn between its parts in different vertebrate groups. Seven homologies are considered: 1, the middle ear cavity/spiracular pouch; 2, the stapes/hyomandibula; 3, the stapedial/hyomandibular processes; 4 the tympanic membrane; 5, the otic notch; 6, the fenestra ovalis; 7, and the stapedial/hyomandibular foramen. The reasons leading to assessments of homology are reviewed. Homologies 1 and 2, based largely on embryological evidence, are fairly (...) robust, though there are arguments about the details. Homologies 3, 4 and 5 stem from ideas about early tetrapod evolution, and were influenced by contingent factors including the order and time of discovery of early fossil taxa, and perceptions of their phylogeny which resulted from this. They were also influenced by ideas of the evolution of terrestriality among tetrapods. Most of the conceptions have been overturned in recent years by new fossil discoveries and new ways of looking at old data. Homology 6 has been little considered. One possible hypothesis, placed in a strictly archetypal theoretical framework has been ignored but deserves consideration on other grounds. Homology 7 depends on how tetrapods are characterised, not a question which has posed difficulties until recently, but which is likely to with the discovery of intermediate fossil forms. (shrink)